PROJECT SUMMARY/ ABSTRACT Developing shorter, safer, more effective drug regimens for active tuberculosis (TB) is a critical public health priority. However, a lack of reliable intermediate clinical trial endpoints constrains accurate regimen selection for Phase 3 trials. Moreover, current surrogate markers cannot explain the root causes of poor outcomes, namely the functional adaptations that enable survival of genetically-susceptible, drug-tolerant M. tuberculosis (Mtb) subpopulations. Our objective is to evaluate sputum Mtb transcriptional profiling as a novel biomarker for predicting relapse and as a surrogate endpoint for clinical trials. Our central hypothesis is that TB treatment outcomes are driven by Mtb physiologic changes measurable via pathogen-targeted transcriptional profiling. Our long-term objective is to develop novel surrogate markers and provide new biologic insights into drug tolerance through direct, in vivo molecular monitoring of Mtb populations during treatment. Our scientific approach will be to perform sputum Mtb transcriptional profiling in culture-confirmed, drug-susceptible pulmonary TB patients co-enrolled in a large, Phase 3, open-label, randomized clinical trial led by the CDC TB Trials Consortium (TBTC) and the NIAID/DAIDS AIDS Clinical Trials Group (ACTG). Study 31/ACTG 5349 will compare two 4-month, high-dose rifapentine-based regimens (one including a fluoroquinolone) with standard 6-month TB treatment. At sites in Kenya, Peru, Uganda, and Vietnam, we will collect RNA-preserved sputum at baseline and throughout treatment from patients at high risk of relapse, including HIV-infected patients, and HIV-uninfected patients with cavitation on baseline chest radiograph. In Aim 1, we will perform genome-wide Mtb transcriptional profiling in protocol-correct patients in each treatment arm to provide a comprehensive roadmap of physiologic and pharmacodynamic effects of TB treatment on the Mtb transcriptome, with biological interpretations of key drug-tolerance pathways. Specifically, we will test hypotheses that transcriptional changes specific to drug mechanism of action can serve as pharmacodynamic markers, as well as distinct hypotheses related to rifamycin and moxifloxacin exposure levels. In Aim 2, we will perform Mtb transcriptional profiling in all culture-/genotype-confirmed relapses and matched controls with relapse-free cure. We will build advanced pharmacokinetic models to select Mtb transcripts that can accurately predict relapse and serve as surrogate endpoints for clinical trials. Aim 2 will also produce an integrative systems pharmacology model to explain between-patient differences in treatment outcomes. Our research program has the potential to inaugurate a new era in which drug-development is based not on culture-based surrogates but on precise, in vivo molecular markers of pathogen physiologic state during TB treatment.